[asterisk-bugs] [JIRA] (ASTERISK-28140) repeated segmentation faults

[George Joseph (JIRA)]

     [ https://issues.asterisk.org/jira/browse/ASTERISK-28140?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

George Joseph updated ASTERISK-28140:
-------------------------------------

    Assignee: Eyal Hasson  (was: Unassigned)
      Status: Waiting for Feedback  (was: Triage)

[~eyal at kolhl.com] I messed up.  Although this issue was triggered by ARI, it's not the issue I was thinking of.  The one I was thinking of involved a stasis message type whose reference count increased over time until it started spitting out the refcount exceeded messages.  That was the one that was just fixed.  Your's actually happened during the process of an audio frame.  Here was the sequence...

* Something happened during the processing of the frame to cause the refcount to exceed 100000.  That's a problem in itself.
* Since that should never happen, asterisk attempted to print a message and call assert().
* The assert() call then attempted to print a backtrace showing the call stack.
* During the generation of the backtrace (which occurs in the libbfd system library) something got corrupted and generated a segfault which crashed asterisk.

If the segfault hadn't happened, asterisk probably would have simply printed the message and backtrace and continued.

So now we have 2 issues...  The refcount > 100000 issue and the segfault during the backtrace generation.

Let me dig into this a bit.

Can you post a few more backtraces though?  I want to see if they're all in the same place.  Just the *thread1.txt file is enough.


> repeated segmentation faults 
> -----------------------------
>
>                 Key: ASTERISK-28140
>                 URL: https://issues.asterisk.org/jira/browse/ASTERISK-28140
>             Project: Asterisk
>          Issue Type: Bug
>      Security Level: None
>          Components: Core/PBX
>    Affects Versions: 13.22.0
>         Environment: CentOS 7, FreeBPX 14.0.4.1, Dell PowerEdge R430
>            Reporter: Eyal Hasson
>            Assignee: Eyal Hasson
>              Labels: pjsip
>         Attachments: core.phonelinuxsrv.kolhl.com-2018-10-30T13-41-03+0200-brief.txt, core.phonelinuxsrv.kolhl.com-2018-10-30T13-41-03+0200-full.txt, core.phonelinuxsrv.kolhl.com-2018-10-30T13-41-03+0200-locks.txt, core.phonelinuxsrv.kolhl.com-2018-10-30T13-41-03+0200-thread1.txt
>
>
> System is running a large ARI application with 1000 lines. It ran smoothly for 4 months. Suddenly yesterday night it started getting segmentation faults, and crashes often - sometimes after 10 minutes, and sometimes after a 3 hours.



--
This message was sent by Atlassian JIRA
(v6.2#6252)